Background PDE2 (phosphodiesterase 2) is upregulated in human heart failure. Cardiac PDE2‐transgenic mice are protected against contractile dysfunction and arrhythmias in heart failure but whether an acute elevation of PDE2 could be of therapeutic value remains elusive. This hypothesis was tested using cardiac PDE2 gene transfer in preclinical models of heart failure. Methods and Results C57BL/6 male mice were injected with serotype 9 adeno‐associated viruses encoding for PDE2A. This led to a ≈10‐fold rise of PDE2A protein levels that affected neither cardiac structure nor function in healthy mice. Two weeks after inoculation with serotype 9 adeno‐associated viruses, mice were implanted with minipumps delivering either NaCl, isoproterenol (60 mg/kg per day), or isoproterenol and phenylephrine (30 mg/kg per day each) for 2 weeks. In mice injected with serotype 9 adeno‐associated viruses encoding for LUC (luciferase), isoproterenol or isoproterenol+phenylephrine infusion induced left ventricular hypertrophy, decreased ejection fraction unveiled by echocardiography, and promoted fibrosis and apoptosis assessed by Masson's trichrome and Tunel, respectively. Furthermore, inotropic responses to isoproterenol of ventricular cardiomyocytes isolated from isoproterenol+phenylephrine‐LUC mice loaded with 1 μmol/L Fura‐2AM and stimulated at 1 Hz to record calcium transients and sarcomere shortening were dampened. Spontaneous calcium waves at the cellular level were promoted as well as ventricular arrhythmias evoked in vivo by catheter‐mediated ventricular pacing after isoproterenol (1.5 mg/kg) and atropine (1 mg/kg) injection. However, increased PDE2A blunted these adverse outcomes evoked by sympathomimetic amines. Conclusions Cardiac gene therapy with PDE2A limits left ventricle remodeling, dysfunction, and arrhythmias evoked by catecholamines, providing evidence that increasing PDE2A activity acutely could prevent progression toward heart failure.

ABSTRACT BACKGROUND Constitutive cardiac PDE2 activation was shown to protect against contractile dysfunction and arrhythmia in heart failure (HF). However, it remains unknown whether an acute elevation of PDE2 is efficient to prevent maladaptive remodeling and arrhythmia. In this study we tested whether increasing acutely PDE2A activity in preclinical models of HF using cardiac PDE2 gene transfer could be of therapeutic value. METHODS AND RESULTS C57BL/6 male mice were injected with serotype 9 adeno-associated viruses (AAV9) encoding for PDE2A, or luciferase (LUC). Cardiac function assessed by echocardiography unveiled neither structural change nor dysfunction consecutive to PDE2A overexpression while AAV9 inoculation led to a ≈10-fold rise of PDE2A protein levels. Two weeks after AAV9 injections, mice were implanted with osmotic minipumps delivering NaCl or isoproterenol (Iso) (60 mg/kg/day) or Iso and phenylephrine (Iso+Phe, 30 mg/kg/day each) for 2 weeks. In LUC mice, chronic infusion with Iso increased left ventricular (LV) weight over body weight ratio, promoted fibrosis and decreased ejection fraction, but animals overexpressing PDE2A were protected towards these deleterious effects. Similarly, concomitant treatment with Iso+Phe promoted LV contractile dysfunction, fibrosis and apoptosis in LUC mice, while PDE2A overexpression limited these adverse outcomes. Furthermore, inotropic responses to Iso of ventricular cardiomyocytes isolated from Iso+Phe-LUC mice loaded with 1 µmol/L Fura-2AM and stimulated at 1 Hz to record calcium transients and sarcomere shortening were dampened. Chronic treatment with catecholamines favoured spontaneous calcium waves upon β-AR stimulation at the cellular level and promoted susceptibility to ventricular arrhythmias in vivo evoked by catheter-mediated ventricular pacing after Iso and atropine injection. However, these adverse effects were blunted by the cardiac gene therapy with PDE2A. CONCLUSION Gene therapy with PDE2A limits cardiac adverse left ventricle remodeling and dysfunction induced by catecholamines as well as ventricular arrhythmias, providing evidence that acutely increasing PDE2A activity could prevent progression towards HF.

Chronic ß-AR activation is detrimental because it promotes cardiac remodeling and leads to heart failure (HF). Multiple cyclic nucleotide phosphodiesterases (PDEs) finely tune ß-AR responses by degrading and compartmentalizing cAMP. PDE2A is upregulated in HF, and PDE2A-transgenic mice are protected against catecholamine-induced arrhythmias. Since chronic treatment with PDE inhibitors increases mortality in HF and PDE2 overexpression seems cardioprotective, we postulated that decreasing cAMP levels by overexpressing PDE2A in the heart using gene therapy may have therapeutic effects. C57BL/6N male mice were injected with serotype 9 adeno-associated viruses (AAV9, 10e12 viral particles) encoding for PDE2A, or luciferase (LUC). Two weeks later, mice were implanted with osmotic pumps infusing NaCl or isoprenaline (Iso) and phenylephrine (IP, 30 mg/kg/day each) for 2 weeks. Cardiac function was assessed by echocardiography, susceptibility to arrhythmias by catheter-mediated ventricular pacing after Iso (1.5 mg/kg) and atropine (1 mg/kg) injection. Ventricular cardiomyocytes were isolated, loaded with 1 μM Fura-2AM and stimulated at 1 Hz to record calcium transients (CaT), sarcomere shortening (SS) and pro-arrhythmogenic spontaneous calcium waves (SCW). In LUC mice, IP treatment increased left ventricular weight over body weight ratio by 35 ± 5.7% (P = 0.0001) and decreased ejection fraction by 29 ± 2.4% (P < 0.0001). Both parameters were improved in animals subjected to AAV9-PDE2A injection (P = 0.039 and P = 0.0003, respectively) increasing by ∼10 fold PDE2A levels. Ventricular tachycardias were triggered in IP-LUC mice (P = 0.016) but not in animals treated with AAV9-PDE2A. In isolated myocytes from control mice, ß-AR stimulation (3 nM Iso) increased SS by 355% and CaT by 86% (P < 0.0001). These inotropic effects were blunted in cells from IP treated animals with SS increasing by 157% (P = 0.0002, vs. LUC-NaCl) and CaT by only 55% (P = 0.06 vs. IP-LUC at baseline). However, the Iso effects were preserved in cardiomyocytes from animals injected with AAV9-PDE2A (SS: P < 0.0001, CaT: P < 0.0001 vs. IP-PDE2A at baseline), which exhibited less SCW upon maximal ß-AR stimulation (Iso, 100 nM, P = 0.02 vs. IP-LUC). Our results suggest that gene therapy with PDE2A limits cardiac hypertrophy and dysfunction induced by catecholamines as well as ventricular arrhythmias.

New Findings What is the central question of this study? Can imipramine, an antidepressant agent that is a cationic amphiphilic drug that interferes with the phosphatidylinositol 4,5‐bisphosphate (PI(4,5)P2) interactions with proteins maintaining the tubular system, be validated as a new detubulating tool? What is the main finding and its importance? Imipramine was validated as a more efficient and less toxic detubulating agent of cardiomyocytes than formamide. New insights are provided on how PI(4,5)P2 is crucial to maintaining T‐tubule attachment to the cell surface and on the cardiotoxic effects of imipramine overdoses. AbstractCardiac T‐tubules are membrane invaginations essential for excitation–contraction coupling (ECC). Imipramine, like other cationic amphiphilic drugs, interferes with phosphatidylinositol 4,5‐bisphosphate (PI(4,5)P2) interactions with proteins maintaining the tubular system connected to the cell surface. Our main purpose was to validate imipramine as a new detubulating agent in cardiomyocytes. Staining adult rat ventricular myocytes (ARVMs) with di‐4‐ANEPPS, we showed that unlike formamide, imipramine induces a complete detubulation with no impact on cell viability. Using the patch‐clamp technique, we observed a ∼40% decrease in cell capacitance after imipramine pretreatment and a reduction of ICa,L amplitude by ∼72%. These parameters were not affected in atrial cells, excluding direct side effects of imipramine. β‐Adrenergic receptor (β‐AR) stimulation of the remaining ICa,L with isoproterenol (Iso) was still effective. ECC was investigated in ARVMs loaded with Fura‐2 and paced at 1 Hz, allowing simultaneous measurement of the Ca2+ transient (CaT) and sarcomere shortening (SS). Amplitude of both CaT and SS was decreased by imipramine and partially restored by Iso. Furthermore, detubulated cells exhibited Ca2+ homeostasis perturbations. Real‐time cAMP variations induced by Iso using a Förster resonance energy transfer biosensor revealed ∼27% decreased cAMP elevation upon β‐AR stimulation. To conclude, we validated a new cardiomyocyte detubulation method using imipramine, which is more efficient and less toxic than formamide. This antidepressant agent induces the hallmark effects of detubulation on ECC and its β‐AR stimulation. Besides, we provide new insights on how an imipramine overdose may affect cardiac function and suggest that PI(4,5)P2 is crucial for maintaining T‐tubule structure.

Université Paris-Saclay, Inserm, Signaling and Cardiovascular Pathophysiology, UMR-S 1